Genetic structure of the (Mustela putorius) and its implication for conservation strategies

C. Pertoldi1,2,3, P. Breyne4, M. T. Cabria5, D. Halfmaerten4, H. A. H. Jansman6, K. Van Den Berge4, A. B. Madsen3 & V. Loeschcke1

1 Department of Ecology and Genetics, University of Aarhus, Ny Munkegade, Aarhus C, Denmark 2 Department of Applied Biology, Estaci ´on Biol ´ ogica Do˜ nana, CSIC, Pabell ´on del Peru´ , Seville, Spain 3 Department of Wildlife Ecology and Biodiversity, National Environmental Research Institute, Rønde, Denmark 4 Institute for Forestry and Game Management, Gaverstraat, Geraardsbergen, Belgium 5 Departamento Zoologia y BCA, Facultad Farmacia, Universidad del Pais Vasco, Vitoria-Gasteiz, Spain 6 Alterra-Centre for Ecosystem Studies, AA Wageningen, The Netherlands Keywords polecat; Mustela putorius; microsatellite DNA; population structure; assignment test; bottleneck; population fragmentation; postglacial recolonization.

Correspondence Cino Pertoldi, Department of Ecology and Genetics, University of Aarhus, Building 540, Ny Munkegade, DK-8000 Aarhus C, Denmark. Fax: +4586127191 Email: [email protected] Received 20 July 2005; accepted 23 November 2005 doi:10.1111/j.1469-7998.2006.00095.x

Abstract During the last century, the European polecat Mustela putorius populations in most of Europe declined and survived in fragmented patches, because of habitat alterations and direct persecution. To assess the genetic consequences of the demographic decline and to describe the spatial pattern of genetic diversity, 250 polecats sampled at seven localities from five European countries – Poland, Denmark (southern Denmark and northern Denmark), Spain, Belgium (eastern and western) and the Netherlands – were screened by means of nine microsatellite loci. Genetic diversity estimated by mean expected heterozygosity (HE) and allelic richness (AR) were moderately high within populations [range: 0.50 (northern Denmark) ≤ HE ≤ 0.64 (Poland) and 1.33 ≤ AR ≤ 7.80] as compared with other carnivores and mustelids. Bottleneck tests suggested that polecat populations in southern Denmark and Poland have declined recently and populations from northern Denmark and the Netherlands have expanded recently, whereas the remaining populations did not show any sign of demographic change. Recent demographic changes could suggest that some of the populations are still not in equilibrium, which could partly explain the relatively high genetic variability observed in polecat populations despite the drastic decline in population size observed in several European countries. A significant heterozygote deficiency [FIS =0.19; 0.01 ≤ 95% confidence interval (CI) ≤ 0.32] suggests substructuring within the total European sample.

Partitioning of the genetic variation among sampling locations (FST =0.14; 0.06 ≤ 95% CI ≤ 0.23) and pairwise FST between localities (range: 0.01 ≤ FST ≤ 0.37) without any correlation with the geographic distances between localities were found, suggesting a recent divergence and a restriction of gene flow between populations and the action of genetic drift. An assignment test showed that the Polish and the northern Danish populations were the most unique, whereas the other populations were partially admixed. Factorial component analysis tests indicate a subdivision of the total sample into two distinct groups: one including the samples from Poland and the two Danish localities and the second group comprising the remaining localities investigated. The observed pattern of genetic differentiation is suggested to be due to two main routes of recolonization after the last glacial period. To compare the results obtained with microsatellite data, the most variable region of the mitochondrial DNA (d-loop) was sequenced and different phylogenetic reconstructions and genetic diversity analyses based on nucleotide (π) and haplotype diversity (h) measures within populations were performed using a subsample of populations. The lack of well-defined geographical structure, as well as the reduced level of mitochondrial DNA variability (p: 0.00274 ± 0.00038; h: 0.876 ± 0.028) that was found, has been previously reported in several studies on different carnivores and supports the hypothesis of post-glacial recolonization from southern or eastern refugees of Europe as suggested by the microsatellite data. Implications for conservation strategies of the polecat at the European level are discussed.

Introduction

Present status of the polecat in Europe The polecat Mustela putorius is a medium-sized carnivore that lives in many parts of Europe (Mitchell-Jones et al., 1999). Its preferred habitat is near bodies of water, like freshwater lakes, rivers and wetlands. Polecats are also found on the edges of forests and grasslands with islands of scrub trees. The polecat is listed in Appendix III of the Bern Convention (European Council, 1979, http://www.eko.org.pl/1kp/prawo_html/konw_bernenska_z3.html) and in Appendix V of the European Community Habitats & Species Directive (1992, http://europa.edu.int/comm/environment/nature/nature_conservation/eu_nature_legislation/habita ts_directive/index_en.htm). The polecat is of some conservation significance as it is considered to be vulnerable in most parts of Europe and threatened in some parts (Birks & Kitchener, 1999). Its distribution range and population densities have decreased in several European countries over the past few decades (Santos Reis, 1983; Vigna Taglianti, 1988; Weber, 1988; Blanco & Gonzalez, 1992; Saint-Girons et al., 1993; Stubbe & Stubbe, 1994; Libois, 1996; Baghli & Verhagen, 2003; Møller et al., 2004). In some other countries however, as for example the Netherlands or in some regions of France, the status of the polecat has been reported as stable (Hollander & Van der Reest, 1994). In Flanders, the northern part of Belgium, polecats never lost the status of ‘commonly spread’ although densities started to decline some decades ago (Van Den Berge & De Pauw, 2003). In Denmark, according to the Danish game bag record the number of polecats killed by hunting during the last 60 years has been declining (Møller et al., 2004). Habitat fragmentation and degradation and more particularly the drainage of wetlands as well as changes in the agricultural landscape have been suggested as the principal causes for the decrease of populations of the polecat in Europe (Blanco & Gonzalez, 1992). Because this mustelid is supposed to be affected by habitat fragmentation and degradation (Bright, 1993), it is considered as an indicator species in human impact studies.

Genetic and demographic consequences of habitat fragmentation In conservation biology there is a growing awareness of the consequences of habitat loss and over-exploitation, which often leads to a reduction in population sizes. Today, habitat fragmentation is considered as one of the most serious threats to the survival of animal populations. This is a result of the spatial structuring of populations caused by the loss of area, the reduction in area of remaining fragments and increased distance between fragments (Gerlach & Musolf, 2000). Habitat fragmentation produces a reduction in genetic diversity because of a loss of variability due to low effective population size (NE) and an increased genetic differentiation between remaining fragments. Population subdivision is therefore expected to reduce the adaptive potential that a population has to face up to environment change (Lacy, 1997; Bijlsma, Bundgaard & Boerema, 2000a). Declining adaptive variation might interact negatively with the demographic consequences of habitat loss, raising the risks of extinction (Bijlsma et al., 2000a). The effect of fitness reduction observed in populations with low NE is thought to be partly the result of increased homozygosity for (partially) recessive deleterious alleles (Charlesworth & Charlesworth, 1987). Several studies have documented inbreeding depression in wild populations (reviewed in Keller & Waller, 2002). The substantial variation in the level of inbreeding depression between populations and their consequences seem to be strongly taxon dependent (Crnokrak & Roff, 1999; Keller & Waller, 2002). Therefore, there is a need for a concerted effort to study patterns of genetic variation in a wide variety of taxa at different spatial scales. Studies of spatial population structure in have revealed that most mammalian populations are genetically subdivided, with relatively small NE and low dispersal rates that are consequences of the social and mating systems. The scale of subdivision varies considerably, suggesting that genetic structure is influenced by complex interactions between social organization, dispersal tendencies and environmental factors (Chepko-Sade & Halpin, 1987). We wanted to assess the likelihood for persistence of the polecat populations studied to estimate the genetic variability within the populations and to describe the patterns of genetic differentiation among the populations. Recent reduction of NE can generate large genetic distances among populations and may drastically reduce the genetic variability within populations in a short time period (Hedrick, 1999). Molecular data based on microsatellite variation were chosen for our investigation as microsatellites are highly informative and are currently used in many studies addressing conservation issues (Kyle, Davis & Strobeck, 2000; Kyle, Robitaille & Strobeck, 2001; Walker et al., 2001; Kyle, Davison & Strobeck, 2003). Due to their high mutation rate (fast evolving), microsatellites show high levels of genetic variation and can potentially reveal barriers to gene-flow (Tautz, 1989). Furthermore, NE is likely to be correlated to the expected heterozygosity (HE) in equilibrium conditions; therefore, inferences about the relative size of NE of the population studied can be made (Pertoldi & Topping, 2004a,b). Because overlapping sets of microsatellites have been used for other investigations on mustelids, we are able to compare the genetic variability at the interspecific level. Another aim of this investigation consisted in testing whether the pattern of genetic variability within localities and the genetic differentiation among localities are the consequence of recent demographic changes or whether they are due to more ancient events. The genetic data were analysed for the occurrence of population bottlenecks and to infer the past demographic history of the polecat populations in Europe (Cornuet & Luikart, 1996). The knowledge of the populations’ past demographic history is also important for the correct interpretation of the observed population structure, as it is well known that animal populations in the Palearctic Region reacted to the Quaternary climatic fluctuations with repeated local extinctions and shifts in distribution range and size (Hewitt, 1999). Demographic bottlenecks in recolonizing propagules may predictably lead to losses of genetic diversity and to interpopulation divergence (Hewitt, 1999). The complex dynamics during late Pleistocene/Holocene make it difficult to disentangle the genetic consequences of natural climatic and habitat changes from the consequences of human-mediated habitat alterations and direct exploitation of natural populations. Therefore, assessment if the demographic changes have occurred recently or before human-mediated habitat alterations is fundamental for drawing conclusions about the consequences of habitat fragmentation. Comparison with mitochondrial DNA (mtDNA) was also performed as it provides a complementary picture of the genetic variability within polecat populations and helps to delineate the phylogeographic structure of polecats across Europe. For this purpose, the most variable region of mtDNA has been sequenced on a subsample of the samples used for the microsatellites investigation. The level of genetic variability has been estimated based on measures of haplotype (h) and nucleotide (π) diversity within polecat populations. The pattern of distribution of pairwise differences of nucleotide sequences or mismatches has been applied to determine the historical demographic events (Rogers & Harpending, 1992). Different approaches of phylogenetic reconstructions were inferred to establish the phylogeographic patterns and intraspecific relationships among polecat populations across Europe.

Materials and methods We collected 250 polecat samples (hair, blood, muscle and liver tissues) from traffic-killed or hunted polecats from seven localities from five countries in Europe between 1997 and 2004 (see Fig. 1): Poland (P; n=24), southern Denmark (S-DK; n=58), northern Denmark (N-DK; n=25), Spain (SP; n=24), eastern Belgium (E-B; n=46), western Belgium (W-B; n=41) and the Netherlands (NL; n=32).

DNA was extracted using standard phenol/chloroform extraction (Sambrook, Fritsch & Maniatis, 1989) or using a modified CTAB extraction method after Milligan (1992). For the DNA isolation from the hair samples, a different protocol was used as hair contains very low DNA quantities (Taberlet et al., 1996). Overnight incubation in digestion buffer with proteinase K was followed by one phenol and one chloroform:isoamylalcohol (24:1) extraction. After precipitation with isopropanol (100%) at 4°C overnight, DNA was dried and dissolved in 50 µL distilled water.

Microsatellites Genotypes of the polecats were determined by the use of microsatellite primers. Four markers (Mvi57, Mvi87, Mvi111 and Mvi114) were developed for Mustela vison (O’Connell, Wright & Farid, 1996), Gg14 was developed for Gulo gulo (Walker et al., 2001) and the last four (Mvi022, Mvi072, Mer009 and Mer041) were developed for Mustela vison and Mustela erminea (Melissa et al., 1999). The microsatellite markers Mvis027 Mvis099 used by Fleming, Ostrander & Cook (1999), Bijlsma et al. (2000b) and Pertoldi et al. (2001) were also tested, but did not show any reliable PCR product or were monomorphic. The PCR reactions for all nine markers were carried out in a 6 µL volume using 0.3U Taq polymerase, 1.5mM MgCl2, 1 µL of diluted DNA, 0.2mM of each dNTP and 0.3 pmol primers. The PCR conditions were 94°C for 3 min, 40 cycles with 94°C for 40 s, 52/60°C for 40 s and 72°C for 40 s. The optimal annealing temperature was determined using a temperature gradient ranging from 50 to 62°C. The amplified loci were analysed on an ABI 373 or ABI 310 automatic sequencer. All alleles were scored manually using the program GENOTYPER 2.5 (PE Applied Biosystems; http://www.perkin-elmer.com/ab/abww0.05.htm). Most of the samples from P, SP, S-DK and N-DK were very small pieces of dried muscle, which created some problems for loci because of the low quality of the DNA, especially for long microsatellites. Therefore, these loci were scored only if the signal of the peaks produced by the alleles was above 65%. Previous studies have shown that allelic dropouts (amplification of just one of two alleles) may occur in the analysis of DNA obtained from degraded tissues, presumably due to a scarcity of intact DNA templates (Zierdt, Hummel & Herrmann, 1996). Therefore, following the procedures describedin Pertoldi et al. (2001, 2005), we extracted DNA three times from the same tissue from 20 individuals in which some of the loci were found to be homozygotic and observed the same genotypes (i.e. the same homozygotic alleles). Hence, the quality and quantity of the DNA extracted and the reproducibility of the results lead us to conclude that problems with allelic drop-outs have not significantly affected our results.

Genetic variability All microsatellite loci and populations were tested for deviations from Hardy–Weinberg expectations (HWE) in FSTAT 2.9.3.2 (Goudet, 2001; http://www.unil.ch/izea/softwares/fstat.html). The program FSTAT was also used to calculate FIS andmean allelic richness (AR) per population (number of alleles corrected for sample size) (ElMousadik, Petit & Pons, 1998). The number of private alleles (PA), HE and the observed heterozygosity (HO) were estimated using the software Pop100gene included in the software Geneclass2 version 2.0 (Piry et al., 2004; http://www.montpellier.inra.fr/CBGPsoftwares/). Significance levels for all the tests performed were adjusted using the sequential Bonferroni correction for multiple comparisons (Rice, 1989).

Analyses of population genetic structure Global and pairwise FST and their associated P-values were estimated using the θ estimator (Weir & Cockerham, 1984) of Wright’s FST (Wright, 1969). The significance levels for the overall and the pairwise values were determined after 10 000 permutations. When random mating could not be assumed within zones, the statistical significance of FST values was evaluated by using exact G-tests by randomizing genotypes among samples not assuming random mating within samples (Goudet et al., 1996). As an alternative measure of the degree of genetic differences among populations, an assignment test was used to determine how unique individual polecats’ genotypes were to the locality from which they were sampled. Individuals were assigned according to the Bayesian method of Rannala & Mountain (1997), with a probability threshold of 5% using the software GeneClass2 version 2.0 (Piry et al., 2004). Genetic differences among individuals and populations were displayed by factorial component analysis (FCA) using Genetix version 4.04 (Belkhir, 2000; http://www.univmontp2.fr/_genetix/genetix/genetix.htm).

Inferring population decline or expansion The occurrence of a population bottleneck or expansion was tested using the software BOTTLENECK 1.2 (Cornuet & Luikart, 1996; http://www.ensam.inra.fr/URLB), assuming an infinite allele model (IAM), a stepwise mutation model (SMM) or a two-phase model of mutation (TPM, with 95% SMM). The Wilcoxon sign-rank test was used to determine whether a population exhibited a significant number of loci with heterozygosity excess under the mutation model. We chose the Wilcoxon test suboption, which has the highest power when the amount of loci screened is less than 20 (Cornuet & Luikart, 1996). The occurrence of a population bottleneck or expansion was tested for each locality separately.

Mitochondrial DNA For the mitochondrial DNA analysis, 20 individuals from Belgium (E-B n=11, W-B n=9), 27 from Denmark (N-DK n=15, S-DK n=12), 15 from Spain and four from Poland were sequenced. To reduce the amount of sequence analysis and as the Netherlands and Belgium are small localities and close together compared with the other regions, both countries were considered as one geographic region. The reason for this limited number of samples is the small amount and bad conservation status of the tissue. The mitochondrial DNA region selected in this study includes the final part of the cytochrome b gene, tRNAPro, tRNAThr, the control region (d-loop) and the initial part of rRNA12S. This fragment of c. 2000 bp length was amplified using the forward primer LutbF (5’-AGAACACCCATTCATCATTATCG- 3’), described in Cabria, Kranz & Go´ mez-Moliner (in press), and the reverse primer LLU12SH91. The standard PCR amplifications were conducted in 15 µL reactions containing 1 µL diluted template DNA, 3.2 pmol of each primer, 1.75mM dNTP, 1.33mM MgCl2, 1.56µL buffer STR 10_ and 0.6U Taq DNA polymerase using the following cycling conditions: an initial denaturing step at 94°C for 5min; 35 cycles of denaturing at 94°C for 50 s, annealing at 58.5°C for 45 s and extending at 72°C for 90 s; and a final extending step of 72°C for 10 min. PCR products were purified by ethanol precipitation, and sequenced directly using the corresponding PCR primers on an ABI- PRISM 3100 DNA sequencer using the dRhodamine Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems) and following the manufacturer’s instructions. Nucleotide sequences were aligned using the default parameters of CLUSTAL X version 1.81 (Thompson et al., 1997) and revised by eye in order to maximize positional homology. Gapped positions were excluded from further mtDNA analysis. The minisatellite repetition of the control region [(TACGCACACG)n in polecats] was also removed from the phylogenetic analysis to reduce ambiguous sites with the selected outgroups. The data set analysed for phylogenetic analysis includes the 24 haplotype nucleotide sequences of the selected mtDNA region from 66 polecats. Four mammalian species were selected as outgroup: Eumetopias jubatus (Steller sea lion; AJ428578), Phoca vitulina (harbor seal; NC_001325), Canis familiaris (dog; NC_002008) and Ursus maritimus (polar bear; AJ428577). Phylogenetic relationships were reconstructed using three methods of phylogenetic inference: minimum evolution (ME) (Rzhetsky & Nei, 1992), maximum parsimony (MP) (Fitch, 1971) and maximum likelihood (ML) (Felsenstein, 1981). The analyses were performed with PAUP* version 4.0b10 (Swofford, 1998), with 10 random addition sequences and TBR branch swapping. MP analyses were executed using a 3:1 transversion (Tv):transition (Ts) ratio. The best-fit model of nucleotide substitution was determined using the program ModelTest version 3.06 (Posada & Crandall, 1998) and selected according to the Akaike information criterion (AIC). The model selected corresponds with general time reversible (GTR) (Tavaré , 1986) with proportion of invariable sites and shape parameter of the gamma distribution (GTR+I+G). The robustness of the recovered trees was assessed with non-parametric bootstrap proportions (BPs- 1000 pseudoreplicates forME and MP, and 100 pseudoreplicates for ML). Alternatively, the NETWORK software version 4.0 (Bandelt, Forster & Röhl, 1999; http://www.fluxus- engineering.com) was used to construct a medianjoining (MJ) network. MJ is a powerful method that reconstructs phylogenies based on intraspecific genetic differentiation (Posada & Crandall, 1998; Bandelt et al., 1999). The data matrix included the nucleotide sequences of the mtDNA region from all the polecats sequenced.

Haplotype diversity, nucleotide diversity and pairwise mismatch distribution The h and π (Nei, 1987) values were estimated using the DnaSP program version 4 (Rozas & Rozas, 1997). These measures were used to define the level of genetic variation within the populations using a data matrix based on the total nucleotide sequences of the mtDNA of individuals from Belgium, Denmark and Spain. The Polish population was discarded from the analysis because of the small number of individuals sequenced (n=4). The distribution of pairwise genetic differences or mismatch distribution was generated using an inference method for the detection of population expansion or decline events (Rogers & Harpending, 1992). All the haplotype nucleotide sequences reported in this investigation are available in GenBank under accession numbers from AY962022 to AY962045.

Results

Microsatellites

Genetic variability The nine microsatellite loci were variable in all the samples (ranging from two to 16 alleles), but with different degrees of variability. HE and AR were moderately high within populations [0.50 (N-DK) ≤ HE ≤ 0.64 (P) and 1.33 ≤ AR ≤ 7.80; see Table 1]. The mean allelic richness was highest for E-B (AR=4.20) and lowest for S-DK (AR=3.51). Overall significant deviations from HWE were found when pooling all the samples from the seven localities [FIS =019; 0.01 ≤ 95% confidence interval (CI) ≤ 0.32; P ≤ 0.0001]. Significant heterozygote deficiency at the locus level was found five times out of the 63 tests for HWE after Bonferroni correction (see Table 1). In particular, locus Mvi87 showed four times significant deviations from HWE in S-DK, E-B, W- B and NL. However, because of the limited number of loci investigated, we decided to include this locus in all the statistical tests. Within localities, significant heterozygote deficiency was found in S-DK (FIS =0.20), E-B (FIS =0.13), W-B (FIS =0.24) and NL (FIS =0.25) (see Table 1). Significant deviation from HWE was also found when excluding locus Mvi87 (results not shown). PAs were found in all the samples investigated with the exception of NL (nPA=13). Some of these alleles occurred at relatively high frequency (frequency range: 0.01 ≤ PA ≤ 0.15; mean - SE=0.04 ± 0.01; median=0.03; see Table 1).

Analysis of the population genetic structure The proportion of correctly assigned individuals in the geographically distinct locations was relatively high (66.8%), ranging from 54 to 79%: P=79%, S-DK=64%, N-DK=76%, SP=54%, E- B=74%, W-B=59% and NL=66%. A significant partitioning of the genetic variation among sampling locations was found (FST=0.14; 0.06 ≤ 95% CI ≤ 0.23). All the pairwise FST values between localities were significant with the exception of the comparisons including SP (ranging from 0.01 ≤ FST ≤ 0.37; see Table 2). The factorial component analysis plot (Fig. 2) shows that polecats sampled from different localities are clustered in two distinct groups, suggesting a genetic differentiation. Polecats from Poland and Denmark are clearly separated from the individuals from Spain, Belgium and the Netherlands.

Inferring population decline or expansion Wilcoxon tests were not significant for E-B, W-B and SP assuming IAM, TPM or SMM (P40 > 05), but showed a significant heterozygote excess in P and S-DK assuming IAM, suggesting that polecat populations have recently declined. The populations from N-DK and NL showed a significant heterozygote deficiency assuming SMM, suggesting a recent expansion (see Table 3).

Mitochondrial DNA

Mitochondrial DNA diversity and inferred phylogeny The mitochondrial DNA region showed 24 haplotypes within the four populations studied (Table 4). The Spanish population is characterized by the highest number of haplotypes (n=10; Mph: 3, 7, 8, 15 and 19–24) and shares one haplotype with the Belgian and Danish populations (Mph2). The Belgian population has five unique haplotypes (Mph: 1, 5, 11, 12, 17) and shares two haplotypes with the Danish population (Mph: 4, 16), which has three unique haplotypes (Mph: 6, 10, 18). The Polish population, in spite of the low number of sequences analysed, also has three unique haplotypes (Mph: 9, 13, 14). The haplotype nucleotide sequences cover a total length of 1534 bp, excluding the minisatellite repetition of the control region. Of all nucleotides, 317 were parsimony informative and 952 were constant. The nucleotide frequencies were A: 32.75%; C: 24.24%; G: 16.01%; T: 27%. Two main clades were recovered with moderately high bootstrap value based on MP and ML (Fig. 3). The topology and the statistical support of the branches obtained byMPand ML did not differ substantially. On the contrary, theME trees recovered fell into unresolved relationships between haplotypes, with the exception of one group representing the Spanish population, which was supported by a high bootstrap value. Clade I was characterized by a low resolution because of the lack of structure and phylogenetic relationships between haplotypes. Only one significant group (subgroup I) constituting the Spanish haplotypes was detected by the three methods of phylogenetic inference (96% ME, 71% MP and 94% ML).

Two haplotypes (Mph2 and Mph17) found in Spain, Denmark and Belgium were clustered in another group (subgroup II) also placed in clade I. Clade II, representing five haplotypes found only in the northern polecat populations, was supported by moderate statistical power. HaplotypesMph10 andMph11 were located in the same group (subgroup 3) with low bootstrap support. The remaining haplotype 1 was placed as the most basal haplotype. The mixture of the haplotypes among the northern and southern populations shows a low genetic differentiation between the polecat populations. To corroborate the low genetic differentiation among the populations, additional intraspecific phylogeny analyses were performed based on networking approaches. The results obtained did not differ substantially from the ones provided by the MJ network. The MJ network showed a relatively high number of mutations among some of the haplotypes (ranging from 1 to 7), and the relationships among haplotypes were in some cases defined by the introduction of missing intermediate haplotypes (Fig. 4). The two clades revealed previously were also well defined by the intraspecific phylogeny and connected by a higher branch, suggesting that the maximum difference is presented in these two main groups. The topology yielded by the MJ network did not fit well with a clearly defined star- like phylogeny and the mixture of haplotypes from different regions also indicates little genetic variation among the different geographic areas analysed. The lack of resolution could be due to the low number of geographic regions studied. The values of π and h for mtDNA, evaluated as the level of genetic variation for each polecat population, are given in Table 4. The global π and h measures for the polecat populations were 0.00274 (SD=0.00038) and 0.876 (SD=0.028), respectively. The overall π and h measures for the Spanish, Belgian and Danish populations were 0.00274 (SD=0.00038) and 0.876 (SD=0.028), respectively. The Spanish population has the highest levels of π and h. Although the estimated h for the Belgian population showed a similar value as the Danish population, higher levels of π were found in the Belgian population. Additionally, similar values of total h estimates were revealed when the nucleotide sequences of the Polish population were included in the analysis (h=0.89_0.026), with an increase of π from 0.00274 to 0.00303 (SD=0.00037). The levels of genetic diversity of the three populations studied are moderately low. This result is consistent with the phylogenetic analyses where no geographical structure was revealed, showing closely related mtDNA lineages between the different geographical areas. The pairwise nucleotide differences ranged from 0 to 10. The observed mismatch distribution for the three polecat populations from Belgium, Denmark and Spain revealed a genetic signature of a population expansion (Luikart et al., 2001; Fig. 5).

Discussion

Genetic variability

The studied European polecat populations were not genetically depauperated. The HE and AR levels (0.50 ≤ HE ≤ 0.64 and 1.33 ≤ AR ≤ 7.80) are in the high end of the range compared with other mustelids. Similar levels of genetic diversity, using an overlapping set of microsatellites, have been observed in British (HE =0.536; Domingo-Roura et al., 2003), American pine (HE =0.66; Kyle et al., 2000), wolverines in North America (HE =0.63; Kyle & Strobeck, 2001) and pine martens from the Netherlands (HE =0.57; Bijlsma et al., 2000b). The high genetic diversity found in the polecat populations was unexpected for different reasons. Populations, which fluctuate in density, are often found to have a low genetic diversity (Caballero, 1994). The polecat populations vary following the density fluctuations in their prey (mainly rodents and anuran; Lodé , 1999), as has been reported in several studies on carnivores (Birks & Kitchener, 1999). The heterogeneity of heterozygosity values and the varying allele numbers also support the idea that population size fluctuations have had some influence on the detected genetic structure as genetic drift is expected to create gaps in within-population distributions of microsatellite allele sizes. In theory, the effects of repeated bottlenecks heavily reduce the expected level of heterozygosity of the population (Motro & Thomson, 1982), but in the populations studied here the level of HE was very high despite the bottleneck-like situation caused by density fluctuations. This could be due to a founder effect where few individuals carrying rare alleles occupy a new patch. These rare alleles would have had the chance to increase in frequency and consequently increase HE, whereas in a large population the opposite occurs, as the rare alleles would have barely increased in frequency.A more or less stable polecat population in Europe could therefore also be the reason for the high-observed genetic diversity. Caution should, however, be taken when discussing density fluctuations in populations. It should be mentioned that the minimum value in density experienced by a fluctuating population needs to be very low before a bottleneck-like situation occurs in the population. Such a founder effect could also partly explain the relatively high frequency of some PA observed in the localities investigated. In our study, polecats from P showed the highest HE (HE =0.64), followed by W-B (HE =0.60), SP and E-B (HE =0.59), S-DK (HE =0.58), NL (HE =0.56) and lastly N-DK (HE =0.50). Such differences could indicate differences in NE of the populations compared, but such kinds of comparisons require collecting polecats in localities of the same size, which is not the case in this investigation. The Spanish population showed (despite the small sample size) a relatively high level of AR and HE, which is in accordance with the glacial refuge theory (Hewitt, 1999). The high AR and HE found in Belgium could originate from post-glacial recolonization of western Europe by polecats from the Iberian Peninsula. The reason for the relatively high HE observed in Poland could be due to the admixed composition of this population due to recolonization from multiple refuges in southern Europe. The relatively low AR and HE found in the two Danish localities (S-DK and N-DK) could be due to a rapid recolonization after the last glaciation followed by restricted gene flow from the adjacent German population. The observed HE values are, however, also unexpected because the two populations that showed a sign of recent bottleneck, P and S-DK, had the highest and an intermediate level of HE, respectively, whereas the two populations that showed a sign of recent expansion, S-DK and NL, had the lowest and the second lowest HE value, respectively. Heterozygosity is, however, expected to be relatively insensitive to reductions in population size, unless the reduction has been very severe or long lasting. Changes in the mean number of alleles should be a more sensitive way of assessing demographic changes (Garant, Dodson & Bernatchez, 2000). A relatively high AR was found in the expanding NL population (AR=3.89) and a relatively low AR was found in the declining S-DK population (AR=3.51); however, the lowest AR value (AR=3.39) was found in the supposedly expanding N-DK population, whereas all the other populations’ AR values lay within this range (see Table 1). Unfortunately, this comparison should also be treated with caution for the reasons mentioned above for the HE comparisons. The significant deviation from HWE within the entire European population is probably due to a Wahlund effect (Wahlund, 1928), which is supported by the significant FST values found between the localities compared and by the relatively large 95% CI of the FIS value (FIS =019; 0.01 ≤ 95% CI ≤ 0.32). The significant heterozygote deficiency found within S-DK (FIS =0.20), E-B (FIS =0.13), W-B (FIS =0.24) and NL (FIS =0.25) could also be due to further genetic substructuring and/or to the different degree of habitat fragmentation within the regions, but the different extension of the localities investigated makes these comparisons difficult. Furthermore, the samples for this study were taken over a period of 8 years. This relatively long period of collection could also have contributed to the observed heterozygote deficiency as collecting individuals over several generations in fluctuating populations could have produced the observed pattern.

Analysis of the population genetic structure

The relatively high FST values found between some localities could be due to the effect of drift, which acted stronger after the interrupted gene flow from the neighbouring populations. Because of habitat fragmentation, populations are considerably diverged from each other and show genetic sub structuring. There are two alternative scenarios for genetic divergence between populations. It could be that the observed differentiation reflects the results of recent fragmentation, with the process of genetic drift only just starting to decrease genetic variability. In that case it can be expected that the genetic divergence between populations will increase in the future. Alternatively, the observed differentiation could reflect historical fragmentation, where populations have diverged from each other in the past and currently exist in a certain equilibrium state. The two hypotheses are not mutually exclusive as the observed genetic structure is probably due to both ancient and recent processes. The fact that PAs have been found in all the localities (with the exception of NL), several with a relatively high frequency, could indicate that the geographic regions have differentiated from each other recently. Even if strong quantitative conclusions about the genetic distance cannot be drawn, as measures of genetic divergence tend to show inverse correlations with levels of population polymorphism (Paetkau et al., 1997), no apparent correlation between FST values and the geographic distances was observed as nearly no genetic differentiation between SP and the three western European populations (E-B, W-B and NL) was found. In contrast, higher FST values were found between those three localities and the two Danish locations (S-DK and N-DK) and P. The phylogeographic structure is supported by the FCA plots showing a clear separation into two well-separated groups. However, such a hypothesis should be investigated using a more comprehensive sample collection including polecats from other European countries and especially from the putative south European glacial refuges. The observed genetic pattern supports the hypothesis of a single recolonization route of Belgium and the Netherlands from the Iberian Peninsula, and a multiple recolonisation route for the Polish and Danish populations followed by a genetic differentiation due to gene-flow restriction and drift of the two Danish localities. The recolonization process should have started not longer than 10 000 years ago. Northern Europe was glaciated and almost the entire central Europe was covered with permafrost until c.. 18 000 years ago. Afterwards, almost all of Europe experienced two extremely dry climatic cycles, respectively 13 000 and 10 000 years ago (the Younger Dryas; Starkel, 1991). During those periods, permafrost and dry steppe habitat conditions were likely adverse to the survival of widespread polecat populations in almost the entire northern and central Europe. Results of the assignment procedures showed that a number of individuals were correctly assigned to their population of origin. However, the high percentage of individuals that was not correctly assigned indicates that some localities are not significantly differentiated from each other, that further genetic substructuring exists among localities and/or that some connection between the separate polecat populations existed during the study period (1997 and 2004), which correspond to c. 3–4 generations. We cannot exclude that some gene flow occurs between geographically adjacent populations, like for example S-DK and N-DK, or

E-B, W-B and NL, but probably the action of drift, which produces genetic divergence between populations, overwhelms the homogenizing migration effect.

Inferring population decline or expansion The tests for recent bottlenecks or expansions result in significant outcomes for some of the populations investigated. These findings reinforce our hypothesis suggesting that the current genetic structure of polecats in Europe is partly explained by the most recent population decline. However, we cannot exclude historical population declines of different order of magnitude, which could have occurred on a time-scale covering thousands of years, and were likely due to late glacial or post-glacial habitat changes as previously discussed. The bottleneck tests by Cornuet & Luikart (1996) can only detect severe and recent declines, which occurred within 0.2 NE to 0.4 NE generations. Furthermore, the patterns of diversification at smaller geographical scales remain to be investigated, as the bottleneck test assumes HWE (Cornuet & Luikart, 1996). This assumption is violated for some of the investigated samples. Lastly, we must bear in mind that the programme is also quite conservative and rarely shows a bottleneck in natural populations, especially if the sample size is limited. Therefore, we cannot exclude the possibility of undetected population decline or expansion in some of the populations investigated.

Phylogenetic analysis and genetic diversity The lack of well-defined geographical structure as a result of low genetic divergence among the polecats’ populations, which has been revealed by the phylogenetic analysis of mtDNA, reinforces the post-glacial colonization hypothesis derived from the microsatellite data analysis and suggests that the recolonization process has been relatively quick. Additionally, the mtDNA analysis revealed the presence of two main groups, clade I sharing Spanish haplotypes and clade II characterized by unique northern haplotypes. This result reinforces our hypothesis of alternative colonization pathways from distinct southern and eastern refuges. Increasing the number of individuals investigated from different geographic areas, mainly from eastern regions, might reinforce this hypothesis and also improve the definition of the network.

Comparisons with other carnivore’s studies suggest moderate levels of genetic variability in the mitochondrial DNA genome of polecats. The mtDNA haplotype diversity found in this study (h: 0.876_0.028) is relatively high, in comparison with, for example, the European mink Mustela lutreola with h=0.469 ± 0.939 (Michaux et al., 2005), the sea Enhydra lutris with h=0.412 (Larson et al., 2002), the Eurasian otter Lutra lutra with h=0.16_0.06 (Ferrando et al., 2004) or the dog Canis familiaris with h=0.76 (Randi et al., 2000). The nucleotide diversity is quite low (π: 0.00274 ± 0.00038) compared with estimates obtained for other carnivores, as for example the sea otter Enhydra lutris with π =0.098 ± 0.029 (Larson et al., 2002), the Eurasian otter Lutra lutra with π =0.0006 (Ferrando et al., 2004), the Steller sea lion Eumetopias jumbatus with π =1.7 (Bickham, Patton & Loughlin, 1996) or the dog Canis familiaris with π =1.30 (Randi et al., 2000). A combination of high levels of h and low values of π could be explained by a recent expansion process after a period with reduced population size (Stamatis et al., 2004). This hypothesis is also supported by the mismatch distribution of the pairwise differences, which revealed a genetic signature of population growth (Rogers & Harpending, 1992; Luikart et al., 2001). A similar mismatch distribution of population growth was reported by Davison et al. (2001) for the polecat, suggesting a post-glacial colonization from different European refugees. In contrast, this study revealed a reduced level of mtDNA variability that was not found previously by Davison et al. (2001).

Conservation strategies for polecats in Europe

This study gives a first description of the genetic structure of polecat populations in Europe, which may provide insight into the likelihood for the persistence of these populations and can contribute to their conservation management. The results presented indicate that the European polecat populations investigated did not suffer a recent severe loss of genetic variability despite the fact that populations like Poland and southern Denmark show signs of a recent bottleneck. Population genetic models predict that if inbreeding depression is primarily due to deleterious recessive alleles, then purging (selection against deleterious recessive alleles) should be a plausible mechanism by which populations get rid of the genetic load. If inbreeding is sudden and extreme, NE is reduced and drift will become stronger relative to selection. Therefore, more random fixation occurs. Consequently, because there are more generations and greater opportunity for selection to act before a given inbreeding coefficient is reached, lower rates of inbreeding are expected to be less deleterious than faster inbreeding for the same total level of inbreeding (Day, Bryant & Meffert, 2003). We can hypothesize that purging has acted more effectively on the slow declining populations, which will have several implications for the management of these populations. Populations that have drastically declined such as S-DK and P or that have suddenly expanded, such as N-DK and NL, should be considered as populations with a higher conservation priority, as compared with the slow declining populations. The risk of a higher genetic load in populations that have suddenly declined is relatively higher. Creating an artificial gene flow between the different populations such as those in this study may be a particularly bad idea because of the potential costs associated with supplementation. Recent studies investigating evolutionary models of the effects of gene flow from interior to peripheral populations have shown that supplementation can hinder adaptation (Kirkpatrick & Barton, 1997) and increase the probability that peripheral populations will go extinct by decreasing fitness in recipient populations. Ecological costs can include introduced diseases or competition between natives and nonnatives. The relatively high genetic differentiation found between some of the localities investigated suggests that outbreeding depression could be of concern (Andersen et al., 2002). On the other hand, migration and associated gene flow between populations influence their survival probability, especially when these populations have become small in number. First, a population’s growth rate may be enhanced by immigrant genomes, reducing the effects of inbreeding and maintaining or even increasing genetic variability (Saccheri & Brakefield, 2002), a kind of genetic rescue effect (Richards, 2000). Second, immigrants have an impact on population size by increasing numerical abundance: the socalled (ecological) rescue effect (Brown & Kodric-Brown, 1977). An increased connectiveness among scattered populations would predictably increase NE and reduce the impact of demographic and genetic stochasticity. However, at the same time it will increase the risk of reintroducing deleterious alleles into a slowly declining population, which were purged from the population but were not purged in populations that underwent strong demographic changes over a relatively short time span. Therefore, the main priority when designing a conservation strategy is to identify the populations that have experienced a severe reduction in population size, because such a reduction can result in inbreeding, loss of genetic variation and fixation of deleterious alleles and thereby decrease the evolutionary potential (Maudet et al., 2002). Once these populations have been identified, it is suggested to keep them isolated from the surrounding populations. When the population is sufficiently isolated such that purged deleterious alleles are not reintroduced by immigration (Keller & Waller, 2002), and is not likely to suffer from inbreeding depression in the near future, a reintroduction plan of polecats from other populations into the recently bottlenecked population could be considered.

Acknowledgements

We are very grateful to the Zoology Department (Faculty of Pharmacy, University of the Basque Country), Javier Lopez de Luzuriaga, Ma Asuncion Gomez Gayubo and Sisco Mañas (Departament de Medi Ambient i Habitatge de la Generalitat de Catalunya) for providing the Spanish polecat samples. This work was supported by grants from the Danish Natural Sciences Research Council (21-01-0526 and 21-03-0125), the Forest & Green division of the Flemish government and the Marie Curies Fellowship of the European Community Host Development program under contract number HPMD-CT-2000-00009 for financial support to Cino Pertoldi.

References Andersen, H.D., Pertoldi, C., Scali, V. & Loeschcke, V. (2002). Intraspecific hybridisation, developmental stability and fitness in Drosophila mercatorum. Evol. Ecol. Res. 4, 603–621.

Baghli, A. & Verhagen, R. (2003). The distribution and status of the polecat Mustela putorius L. in Luxembourg. Mamm. Rev. 33, 57–68.

Bandelt, H.J., Forster, P. & Röhl, A. (1999). Median-joining network for inferring intraspecific phylogenies. Mol. Biol. Evol. 16, 37–48.

Belkhir, K. (2000). GENETIX, logiciel sous Windows pour la génétique des populations. Laboratoire Gé nomes et Populations, Université de Montpellier, Montpellier, France. Available from http://www.univ-montp2.fr/_genetix/genetix/genetix.htm.

Bickham, J.W., Patton, J.C. & Loughlin, T.R. (1996). High variability for control-region sequences in a marine : implications for conservation and biogeography of Steller sea lions Eumetopias jubatus. J. Mammal. 77, 95–108.

Bijlsma, R., Bundgaard, J. & Boerema, A. (2000a). Does inbreeding affect the extinction risk of small populations?: predictions from Drosophila. J. Evol. Biol. 13, 502–514.

Bijlsma, R., van de Vliet, M., Pertoldi, C., Apeldoorn, C. & van Zande, L. (2000b). Microsatellite primers from the Eurasian , meles. Mol. Ecol. 9, 2216–2217.

Birks, J.D.S. & Kitchener, A.C. (1999). Ecology of the polecat in lowland England. In The distribution and status of the polecat mustela putorius in Britain in the 1990s: 111–130.

Birks, J.D.S. & Kitchener, A.C. (Eds). London: The Vincent Wildlife Trust.

Blanco, J.C. & Gonzalez, J.L. (1992). Libro Rojo de Los Vertebrados de España. Madrid: ICONA. Bright, P.W. (1993). Habitat fragmentation – problems and predictions for British mammals. Mamm. Rev. 23, 101–111.

Brown, J. & Kodric-Brown, A. (1977). Turnover rates in insular biogeography: effect of immigration and extinction. Ecology 58, 445–449.

Caballero, A. (1994). Developments in the prediction of effective population size. Heredity 3, 657–679.

Cabria, M.T., Kranz, A. & Go´ mez-Moliner, B.J. (in press). Variabilidad genética de la poblacion occidental de vison europeo (Mustela lutreola L., 1761) basada en la secuenciacion de la region control del ADN mitocondrial. Libro de actas del congreso sobre la conservacio n del vison europeo. CL, in press.

Charlesworth, D. & Charlesworth, B. (1987). Inbreeding depression and its evolutionary consequences. Annu. Rev. Ecol. Syst. 18, 237–268.

Chepko-Sade, B.D. & Halpin, Z.T. (1987). Mammalian dispersal patterns: the effects of social structure on population genetics. Chicago: University of Chicago Press.

Cornuet, J.M. & Luikart, G. (1996). Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144, 2001–2014.

Crnokrak, P. & Roff, D.A. (1999). Inbreeding depression in the wild. Heredity 83, 260–270.

Davison, A., Birks, J.D., Brookes, R.C.,Messenger, J.E. & Griffiths, H.I. (2001). Mitochondrial phylogeography and population history of pinemartensMartes martes compared with polecats Mustela putorius. Mol. Ecol. 10, 2479–2488.

Day, S.B., Bryant, E.H. & Meffert, L.M. (2003). The influence of variable rates of inbreeding on fitness, environmental responsiveness, and evolutionary potential. Evolution 57, 1314–1324.

Domingo-Roura, X., Macdonald, D.W., Roy, M.S., Marmi, J., Terradas, J., Woodroffe, R., Burke, T. & Wayne, R.K. (2003). Confirmation of low genetic diversity and multiple breeding females in a social group of Eurasian badgers from microsatellite and field data. Mol. Ecol. 12, 533–539.

El Mousadik, A., Petit, R. & Pons, O. (1998). Identifying populations for conservation on the basis of genetic markers. Conserv. Biol. 12, 844–855.

Felsenstein, J. (1981). Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol. 17, 368–376.

Ferrando, A., Ponsà , M., Marmi, J. & Domingo-Roura, X. (2004). Eurasian , Lutra lutra, have a dominant mtDNA haplotype from the Iberian Peninsula to Scandinavia. J. Hered. 95, 430– 435.

Fitch, W.M. (1971). Toward defining the course of evolution: minimal change for a specific tree topology. Syst. Zool. 20, 406–416.

Fleming, M., Ostrander, E. & Cook, J. (1999). Microsatellite markers for American mink Mustela vison and ermine Mustela erminea. Mol. Ecol. 8, 1352–1354.

Garant, D., Dodson, J.J. & Bernatchez, L. (2000). Ecological determinants and temporal stability of the within-river population structure in Atlantic salmon Salmo salar L. Mol. Ecol. 9, 615–628.

Gerlach, G. & Musolf, K. (2000). Fragmentation of landscape as a cause for genetic subdivision in bank voles. Conserv. Biol. 14, 1066–1074.

Goudet, J. (2001). FSTAT, a program to estimate and test gene diversities and fixation indices version 2.9.3, updated from Goudet 1995. Available from http://www.unilch/izea/softwares/fstat.html.

Goudet, J., Raymond, M., DeMeeus, T. & Rousset, F. (1996). Testing differentiation in diploid populations. Genetics 144, 1933–1940.

Hedrick, P.W. (1999). Highly variable loci and their interpretation in evolution and conservation. Evolution 53, 313–318.

Hewitt, G.M. (1999). Post-glacial recolonization of European biota. Biol. J. Linn. Soc. 68, 1–2.

Hollander, H. & Van der Reest, P. (1994). Red data book of threatened mammals in the Netherlands. Utrecht: Vereniging voor Zoogdierkunde en Zoogdierbescherming. Keller, L.F. & Waller, D.M. (2002). Inbreeding effects in wild populations. TREE 17, 230–241.

Kirkpatrick, M. & Barton, N.H. (1997). Evolution of a species’ range. Am. Nat. 150, 1–23.

Kyle, C., Davis, C. & Strobeck, C. (2000). Microsatellite analysis of North American pine Martes Americana populations from the Yukon and northwest territories. Can. J. Zool. 78, 1150– 1157.

Kyle, C., Davison, A. & Strobeck, C. (2003). Genetic structure of European pine martens Martes martes, and evidence for introgression with M. americana in England. Conserv. Genet. 4, 179– 188.

Kyle, C., Robitaille, J. & Strobeck, C. (2001). Genetic variation and structure of Fisher Martes pennanti populations across North America. Mol. Ecol. 10, 2341–2347.

Kyle, C. & Strobeck, C. (2001). Genetic structure of the American wolverine Gulo gulo populations. Mol. Ecol. 10, 337–347.

Lacy, R.C. (1997). Importance of genetic variation to the viability of mammalian populations. J. Mamm. 78, 320–335.

Larson, S., Jameson, R., Bodkin, J., Staedler, M. & Bentzen, P. (2002). Microsatellite DNA and mitochondrial DNA variation in remnant and translocated sea otter (Enhydra lutris) populations. J. Mammal. 83, 893–906.

Libois, R. (1996). The current situation of wild mammals in Belgium: an outline. Hystrix 8, 35– 41. Lodé , T. (1999). Time budget as related to feeding tactics of European polecat Mustela putorius. Behav. Process. 47, 11–18.

Luikart, G., Gielly, L., Excoffier, L., Vigne, J.D., Bouvent, J. & Taberlet, P. (2001). Multiple maternal origins and weak phylogeographic structure in domestic goats. Proc. Natl. Acad. Sci. USA 98, 5927–5932.

Maudet, C., Bassano, B., Breitenmoser-Wü rsten, C., Gauthier, D., Obexer-Ruff, G.,Michallet, J., Taberlet, P. & Luikart, G. (2002).Microsatellite DNA and recent statistical methods in wildlife conservation management: applications in Alpine ibex (Capra ibex (ibex)). Mol. Ecol. 11, 421– 436.

Melissa, A., Fleming, E., Ostrander, A. & Cook, J.A. (1999). Microsatellites markers for American mink Mustela vison and ermine Mustela erminea. Mol. Ecol. 8, 1352–1354.

Michaux, J.R., Hardy, O.J., Justy, F., Fournier, P., Kranz, A., Cabria, M., Davison, A., Rosoux, R. & Libois, R. (2005). Conservation genetics and population history of the threatened European mink Mustela lutreola, with an emphasis on the West European population. Mol. Ecol., 14, 2373–2388.

Milligan, B. (1992). Plant DNA isolation. In Molecular genetic analysis of populations: a practical approach: 59–88. Hoelzel, A.R. (Ed.). Oxford: IRL Press.

Mitchell-Jones, A.J., Amori, G., Bogdanowicz, W., Krystufek, B., Reijnders, P.J.H., Spitzenberger, F., Stubbe, M., Thissen, J.B.M., Vohralik, V. & Zima, J. (1999). The atlas of European mammals. London, UK: Academic Press/T & AD Poyser Ltd.

Møller, T.B., Pertoldi, C., Madsen, A.B., Asferg, T., Frydenberg, J., Hammershøj, M. & Loeschcke, V. (2004). Genetic variability in the Danish polecats as assessed by microsatellites. Wildl. Biol. 10, 51–59.

Motro, U. & Thomson, G. (1982). On heterozygosity and the effective size of populations subject to size changes. Evolution 36, 1059–1066.

Nei, M. (1987). Molecular evolutionary genetics. New York: Columbia University Press. O’Connell, M., Wright, J.M. & Farid, A. (1996). Development of PCR primers for nine polymorphic American mink Mustela vision microsatellite loci. Mol. Ecol. 5, 311–312.

Paetkau, D., Waits, L., Clarkson, P., Craighead, L. & Strobeck, C. (1997). An empirical evaluation of genetic distance statistics using microsatellite data from bear (Ursidae) populations. Genetics 147, 1943–1957.

Pertoldi, C., Hansen, M.M., Loeschcke, V., Madsen, A.B., Jacobsen, L. & Baagoe, H. (2001). Genetic consequences of population decline in European otter Lutra lutra: an assessment of microsatellite DNA variation in Danish otters from 1883 to 1993. Proc. Roy. Soc. Lond. Ser. B. 268, 1775–1781.

Pertoldi, C., Loeschcke, V., Randi, E., Madsen, A.B., Hansen, M.M., Bijlsma, R. & Van De Zande, L. (2005). Present and past microsatellite variation and assessment of genetic structure in Eurasian badger Meles meles in Denmark. J. Zool. (Lond.) 265, 387–394.

Pertoldi, C. & Topping, C. (2004a). The use of agent-based modelling of genetics in conservation genetics studies. J. Nat. Conser. 12, 111–120.

Pertoldi, C. & Topping, C. (2004b). Impact assessment predicted by means of genetic agent- based modelling. Cri. Rev. Toxicol. 34, 487–498.

Piry, S., Alapetite, A., Cornuet, J.M., Paetkau, D., Baudouin, L. & Estoup, A. (2004). GeneClass2: a software for genetic assignment and first generation migrants detection. J. Hered. 95, 536–539.

Posada, D. & Crandall, K.A. (1998). Model-test: testing the model of DNA substitution. Bioinformatics 14, 817–818.

Randi, E., Lucchini, V., Christensen, M.F., Mucci, N., Funk, S.M., Dolf, G. & Loeschcke, V. (2000). Mitochondrial DNA variability in Italian and east European wolves: detecting the consequences of small population size and hybridisation. Conserv. Biol. 14, 464–473.

Rannala, B. & Mountain, J.L. (1997). Detecting immigration by using multilocus genotypes. Proc. Natl. Acad. Sci. USA 94, 9197–9201.

Rice, W.R. (1989). Analysing tables of statistical tests. Evolution 43, 223–225.

Richards, C. (2000). Inbreeding depression and genetic rescue in a plant metapopulation. Am. Nat. 155, 383–394.

Rogers, A.R. & Harpending, H. (1992). Population growth makes waves in the distribution of pairwise genetic differences. Mol. Biol. Evol. 9, 552–569.

Rozas, J. & Rozas, R. (1997). DnaSP version 2.0: a novel software package for extensive molecular population genetics analysis. Comput. Appl. Biosci. 13, 307–311.

Rzhetsky, A. & Nei, M. (1992). A simple method for estimating and testing minimum-evolution trees. Mol. Biol. Evol. 9, 945–967.

Saccheri, I. & Brakefield, P. (2002). Rapid spread of immigrant genomes into inbred populations. Proc. Roy. Soc. Lond. Ser. B. 269, 1073–1078.

Saint-Girons, M.-C., Maurin, H., Rosoux, R. & Keith, P. (1993). Les mammifères d’eau douce, leur vie, leurs relations avec l’homme. Ministe` re de l’Environnement, Ministère de l’Agriculture et de la Pê che et la Société Française pour l’Etude et la Protection des Mammifères, Paris.

Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989). Molecular cloning. 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

Santos Reis, M. (1983). Status and distribution of the Portuguese mustelids. Acta Zool. Fenn. 174, 213–216.

Stamatis, C., Triantafyllidis, A., Moutou, A. & Mamuris, Z. (2004). Mitochondrial DNA variation in Northeast Atlantic and Mediterranean populations of Norway lobster, Nephrops norvegicus. Mol. Ecol. 13, 1377–1390.

Starkel, L. (1991). Environmental changes at the Younger Dryas – preboreal transition and during the early Holocene: some distinctive aspects in central Europe. Holocene 1, 234–242.

Stubbe, M. & Stubbe, A. (1994). Säugetierarten und deren feldökologische Erforschung im östlichen Deutschland. Tiere Im Konflikt, Vol. 3. Halle-Wittenberg: Martin-Luther- Universitä t. Swofford, D.L. (1998). PAUP_: phylogenetic analysis using parsimony (_and other methods), version 4.0. Sunderland, MA: Sinauer Associates, Inc.

Taberlet, P., Griffin, S., Goosens, B., Questiau, S., Manceau, V., Escaravage, N., Waits, L. & Bouvet, J. (1996). Reliable genotyping of samples with very low DNA quantities using PCR. Nucleic Acids Res. 24, 3189–3194.

Tautz, D. (1989). Hypervariability of simple sequences as a general source for polymorphic DNA markers. Nucleic Acids Res. 17, 6463–6471.

Tavaré , S. (1986). Some probabilistic and statistical problems in the analysis of DNA sequences. In Some mathematical questions in biology - DNA sequence analysis: 57–86.

Miura, R.M. (Ed.). American Mathematical Society: Providence, RI. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, J. & Higgins, D.G. (1997). The CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882.

Van Den Berge, K. & De Pauw, W. (2003). Bunzing Mustela putorius Linnaeus, 1758. In Zoogdieren in Vlaanderen – Ecologie en verspreiding van 1987 tot 2002: 323–328. Verkem, S., De Maeseneer, J., Vandendriessche, B., Verbeylen, G. & Yskout, S. (Eds). België : Natuurpunt Studie & JNM Zoogdierenwerkgroep, Mechelen & Gent.

Vigna Taglianti, A. (1988). Stato attuale delle conoscenze sulla biologica e la conservazione dei Carnivori in Italia. Supplemento Alle Ricerche Di Biologia Della Selvaggina XIV, 401–417.

Wahlund, S. (1928). Composition of populations from the perspective of the theory of heredity. Hereditas 11, 65–105.

Walker, C., Vilà, C., Landa, A., Lindén, M. & Ellegren, H. (2001). Genetic variation and population structure in Scandinavian wolverine Gulo gulo populations. Mol. Ecol. 10, 53–63.

Weber, D. (1988). Die aktuelle Verbreitung des Iltisses Mustela putorius in der Schweiz. Rev. Suisse Zool. 95, 1041–1056.

Weir, B. & Cockerham, C. (1984). Estimating F-statistics for the analysis of population structure. Evolution 38, 1358–1370.

Wright, S. (1969). The theory of gene frequencies. Chicago: University of Chicago Press.

Zierdt, H., Hummel, S. & Herrmann, B. (1996). Amplification of human short tandem repeats from medieval teeth and bone samples. Hum. Biol. 68, 185–199.